Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
An Overview1
Gregory F. Gruska
Maureen S. Heaphy
Continued on page 10
COMPLETELY RANDOMIZED DESIGN The block may consist of observations taken at nearly
A completely randomized design is appropriate when the same time or place. If a machine can test four items at
the effects of only one signal variable are being investi- one time, then each run may be regarded as a block of
gated. The effects of any and all noise variables will be four units, each item being a unit.
controlled by either holding constant or randomization. A variety of especially advantageous configurations of
Suppose a total of N experimental units are available block designs have been developed. They are named and
for the experiment and there are k factor levels to be classified by their structure into randomized blocks,
investigated. Then the total N units are assigned randomly incomplete blocks, Latin squares Youden squares, etc.
to the k levels. The sample size in each cell does not Sometimes the factor and the block are of almost equal
need to be equal. interest. In this case a ‘block design’ is almost a ‘two-
This simple one-factor design is called “completely factor experiment,’ but the experimenter must be sure
randomized” to distinguish it from other experiment that the two factors do not interact before using a block
designs where the principle of “blocking” or planned design. If interaction between factors exists or is sus-
grouping has been made part of the structure. pected, the design and analysis for a factorial experiment
APPLICABILITY - The plan is simple and may be the with two factors must be used. In other words a block
best choice when the experimental material is homo- design (with one-way blocking) can be considered as a
geneous and when background conditions can be well one-factor design or a two-factor-no interaction design.
controlled during the experiment. This type of design is The simplest design with one-way block is the
used to estimate and compare the effects of one ‘Randomized Block Design.’ This and others will be
independent variable, the signal. In some cases the discussed.
precision of that estimate can be quantified; i.e., how
much variation is there in the estimate of the effect of the RANDOMIZED BLOCK DESIGN
independent variable. In comparing a number of factor levels, it is clearly
ADVANTAGES - There are three main advantages to a desirable that all other conditions be kept as nearly con-
completely randomized design: stant as possible. Often the required number of tests is
• With only one independent variable, the model and too large to be carried out under similar conditions. In
the calculations are simple and straightforward. such cases, we may be able to divide the experiment into
• Equal sample sizes are not required for each cell. blocks, or planned homogeneous groups. When each
• It allows for the maximum degrees of freedom in such group in the experiment contains exactly one obser-
estimating the error. vation on every treatment, the experimental plan is called
DISADVANTAGES - There are four main disadvantages a randomized block design.
to a completely randomized design: APPLICABILITY - This design recognizes one signal
• Appropriate only when experimental material is and one noise variable. For example, a testing scheme
homogeneous. may take several days to complete. If we expect some
• Restricted to evaluating only one signal factor. measurable differences between days, then a day would
• Appropriate only when the noise conditions can be represent a block. In another situation, several persons
controlled. may be conducting the tests or making the observations,
• Large sample size may be required to obtain suffi- and differences between operators are expected. The
cient precision. tests or observations made by a given operator can be
considered to represent a block. Also, the size of a block
BLOCK DESIGNS may be restricted by physical considerations.
An important class of experimental designs is charac- This type of design is used to estimate and compare
terized by planned grouping. This class is called block factor level effects after removing block effects. Although
designs. The use of blocking arose in comparative experi- the primary interest is to estimate factor effects, it is also
ments in agricultural research, in recognition of the fact possible to estimate the block effects.
that plots that were close together in a field were usually ADVANTAGES - There are three main advantages of a
more alike than plots that were far apart. In industrial and randomized block design:
engineering research, the tool of planned grouping can • The analysis of the data is relatively simple and
be used to take advantage of naturally homogeneous straightforward.
groupings in materials, machines, time, etc., and also to • The experimental material can be divided into
take account of ‘background variables’ which are not blocks or homogeneous groups.
directly ‘factors’ in the experiment. Continued on page 11
CAR I II III IV
ADVANTAGES - The primary advantage of a Latin
FR A B C D square design is the ability to block out the effects of two
Position FL A B C D noise variables thereby isolating the influence of the
RR A B C D primary factor. With this type of design it is possible to
RL A B C D estimate and compare the effects of the two blocked
variables. Also, the analysis is considered to be relatively
One Brand Assigned to a Car simple.
DISADVANTAGES - One disadvantage is that the
design assumes no interactions exist among the three
With this assignment, the test results cannot be variables. A second item, that initially seems to be a
attributed to tire brand differences only since there might restriction, is there must be the same number of levels for
be driver/vehicle influence. To eliminate the driver/ all three variables. If this were not the case, then an
vehicle influence, one might decide to randomly assign incomplete Latin square, called a Youden square, would
the sixteen tires (4 tires of each of 4 brands) to the be appropriate.
four vehicles. The results-of such and assignment could
be as shown as follows. Continued on page 12